A meta study of reduction of speckle noise adopting different filtering techniques

Reduction of noise in digital images is a challenging task for the researchers in digital image processing. Speckle noise creates most critical disturbances and affects the quality of synthetic aperture radar, ultrasound and medical images in the field of biomedical imaging. SAR imagery uses the microwave radiation, which receives coherent summation of reflected radiation. Speckle noise in SAR produces degraded, low quality image which is causing misdirection of active radar in automatic target detection and recognition due to image interpretation. Ultrasound images are affected by speckle noise because the images are obtained with a simple linear scan, ultrasound probe which are produced granular appearance noise. Medical images are also corrupted through the speckle noise during its acquisition and transmission. Image with speckle reduce the image contrast and resolution, thereby decreasing the diagnostic values of the ultrasound image. There are several filtering techniques available to reduce this speckle noise. So, this framework includes several filtering techniques like mean, median, Weiner, log and Gaussian to evaluate the filter performance through the parameters like mean square error and peak signal to noise ratio to attenuate the speckle noise without degrading the image information from the synthetic aperture radar, ultrasound and medical imagery. Experimental analysis has been done under MATLAB environment. MSE, PSNR have been calculated and compared for different value of variance for better performance measurement.